Method and system for improved classification of constituent materials

Patent metadata

Field	Value
Publication number	US-2017371010-A1
Application number	US-201515532680-A
Country	US
Kind code	A1
Filing date	Dec 2, 2015
Priority date	Dec 4, 2014
Publication date	Dec 28, 2017
Grant date	—

Abstract

An imaging system and method are disclosed. An MR image and measured B0 field map of a target volume in a subject are reconstructed, where the MR image includes one or more bright and/or dark regions. One or more distinctive constituent materials corresponding to the bright regions are identified. Each dark region is iteratively labeled as one or more ambiguous constituent materials. Susceptibility values corresponding to each distinctive and iteratively labeled ambiguous constituent material is assigned. A simulated B0 field map is iteratively generated based on the assigned susceptibility values. A similarity metric is determined between the measured and simulated B0 field maps. Constituent materials are identified in the dark regions based on the similarity metric to ascertain corresponding susceptibility values. The MRI data is corrected based on the assigned and ascertained susceptibility values. A diagnostic assessment of the target volume is determined based on the corrected MRI data.

First claim

We claim:

1 . A method for imaging a subject, comprising:
reconstructing a magnetic resonance image and a measured B0 field map corresponding to a target volume in the subject using magnetic resonance imaging data acquired from the target volume, wherein the magnetic resonance image comprises one or more bright regions, one or more dark regions, or a combination thereof; identifying one or more distinctive constituent materials corresponding to each of the one or more bright regions in the magnetic resonance image; iteratively labeling each of the one or more dark regions in the magnetic resonance image as one or more ambiguous constituent materials; assigning susceptibility values corresponding to each of the one or more distinctive constituent materials and the one or more iteratively labeled ambiguous constituent materials; iteratively generating a simulated B0 field map based on the assigned susceptibility values corresponding to each of the one or more distinctive constituent materials and the one or more iteratively labeled ambiguous constituent materials; determining a similarity metric between the measured B0 field map and the simulated B0 field map during each iteration; identifying the one or more ambiguous constituent materials in each of the one or more dark regions based on the determined similarity metric to ascertain corresponding susceptibility values; correcting the magnetic resonance imaging data based on the assigned susceptibility values corresponding to the one or more distinctive constituent materials and the ascertained susceptibility values corresponding to the one or more ambiguous constituent materials identified in the magnetic resonance image; and determining a diagnostic assessment of the target volume based on the corrected magnetic resonance imaging data.

2 . The method of claim 1 , wherein reconstructing the magnetic resonance image and the measured B0 field map comprises:
generating the magnetic resonance image using one or more spin echo imaging pulse sequences, one or more gradient echo sequences, or a combination thereof; and generating the measured B0 field map using one or more multi-echo acquisition pulse sequences.

3 . The method of claim 1 , wherein identifying the one or more distinctive constituent materials comprises processing the magnetic resonance image via image segmentation, organ stratification, organ context identification, phase-fields based classification, active contours based classification, level-sets based classification, thresholding-based classification, or combinations thereof.

4 . The method of claim 1 , wherein iteratively labeling each of the one or more dark regions comprises providing an initial assessment of the one or more dark regions based on iteratively comparing corresponding shapes to one or more dipole pattern templates associated with to the one or more ambiguous constituent materials.

5 . The method of claim 4 , wherein iteratively comparing the corresponding shapes to the one or more dipole pattern templates comprises matching the corresponding shapes of the one or more dark regions with the one or more dipole pattern templates using a rigid transformation, an affine transformation, a non-rigid transformation, or combinations thereof.

6 . The method of claim 4 , further comprising receiving the one or more dipole pattern templates corresponding to one or more metal objects, one or more sizes corresponding to the one or more metal objects, or a combination thereof.

7 . The method of claim 1 , wherein iteratively labeling each of the one or more dark regions in the magnetic resonance image comprises labeling the one or more dark regions in the magnetic resonance image with one or more types corresponding to the one or more ambiguous constituent materials, one or more sizes corresponding to the one or more ambiguous constituent materials, or a combination thereof.

8 . The method of claim 1 , wherein iteratively labeling each of the one or more dark regions in the magnetic resonance image comprises iteratively selecting the one or more ambiguous constituent materials from a plurality of ambiguous constituent materials.

9 . The method of claim 1 , wherein assigning the susceptibility values corresponding to each of the one or more distinctive constituent materials and the one or more ambiguous constituent materials comprises using a lookup table, and wherein the lookup table comprises stored correlations between the one or more distinctive constituent materials, the one or more ambiguous constituent materials, or a combination thereof, and corresponding susceptibility values.

10 . The method of claim 1 , further comprising modeling the iterative labeling of each of the one or more dark regions and assigning susceptibility values corresponding to each of the one or more distinctive constituent materials and the one or more iteratively labeled ambiguous constituent materials as an inverse problem.

11 . The method of claim 10 , further comprising:
generating one or more phase magnetic resonance images from the magnetic resonance imaging data, the magnetic resonance image, or a combination thereof; pre-processing the one or more phase magnetic resonance images to remove a corresponding contribution of global phase information while retaining local susceptibility phase changes; and computing a susceptibility matrix, wherein each element in the susceptibility matrix corresponds to a determined susceptibility in a direction of change in a local magnetic field that induces magnetization in the target volume with respect to an externally applied B0 magnetic field; and determining a solution of the inverse problem based on the induced magnetization determined from the one or more pre-processed phase magnetic resonance images, a magnetization caused by the externally applied B0 magnetic field, and the computed susceptibility matrix.

12 . The method of claim 1 , wherein iteratively generating the simulated B0 field map based on the assigned susceptibility values comprises:
generating a susceptibility distribution map corresponding to the magnetic resonance image based on the assigned susceptibility values corresponding to each of the one or more distinctive constituent materials and the one or more iteratively labeled ambiguous constituent materials; and generating the simulated B0 field map from the susceptibility distribution map using a slice-by-slice Fourier domain transformation.

13 . The method of claim 1 , wherein the similarity metric comprises a normalized correlation, a semantic similarity ensemble, or a combination thereof.

14 . The method of claim 1 , wherein identifying the one or more ambiguous constituent materials in each of the one or more dark regions comprises:
localizing a metal object located in the target volume; and determining one or more of a size and a type of the localized metal object.

15 . The method of claim 14 , further comprising automatically initiating specialized imaging pulse sequences to image the one or more dark regions in the target volume proximal the metal object.

16 . The method of claim 1 , further comprising identifying one or more constituent materials in at least one dark region selected from the one or more dark regions based on the one or more ambiguous constituent materials identified in the one or more dark regions adjacent to the at least one selected dark region.

17 . The method of claim 1 , wherein identifying the one or more ambiguous constituent materials comprises identifying one or more of air, bone, and a metal ob

Assignees

• Gen Electric

Inventors

• Shanbhag Dattesh Dayanand
• Mullick Rakesh
• Thiruvenkadam Sheshadri
• Wiesinger Florian
• Chatterjee Sudhanya
• Koch Kevin Matthew

Classifications

• G01R33/4816

  NMR imaging of samples with ultrashort relaxation times such as solid samples, e.g. MRI using ultrashort TE [UTE], single point imaging, constant time imaging · CPC title

• G01R33/243 Primary

  Spatial mapping of the polarizing magnetic field · CPC title

• G01R33/56536 Primary

  due to magnetic susceptibility variations · CPC title

• A61B5/0035

  adapted for acquisition of images from more than one imaging mode, e.g. combining MRI and optical tomography · CPC title

• G01R33/5608

  Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels (image data processing or generation, in general G06T) · CPC title